A Novel Branch and Bound Pure Integer Programming Phase Unwrapping Algorithm for Dual-Baseline InSAR

Abstract

Phase unwrapping (PU) is an important bottleneck restricting the practical application of the interferometric synthetic aperture radar (InSAR) technique. In view of the similarity between solving the ambiguity number of integral cycles in PU for dual-baseline InSAR and pure integer programming (PIP) problem in science of overall planning, a new branch and bound PIP-PU algorithm for dual-baseline InSAR is proposed. A PIP-PU model with the intercept on the vertical axis as the objective function and a ray as the constraint condition is first constructed. Then, how to solve the ambiguity number is given in detail by graphical means. Finally, the axis symmetry theory is introduced to further improve PU efficiency. The proposed algorithm has the advantages of better unwrapping ability even in phase under-sampling areas and abrupt topographic change areas and lower requirement of the baselines. Through two sets of simulated data and one set of real data experiments, the feasibility, effectiveness, and practicability of this proposed algorithm are verified, respectively. In addition, compared with the branch-cut method, quality-guided method, least square method, and minimum cost flow method, the proposed method has the highest accuracy and suboptimal unwrapping efficiency.